US 11,928,513 B1
Cloud affinity based on evaluation of static and dynamic workload characteristics
Peng Hui Jiang, Beijing (CN); Dong Hui Liu, Beijing (CN); Jia Tian Zhong, Beijing (CN); Xing Xing Shen, Beijing (CN); Jia Yu, Beijing (CN); and Yong Yin, Beijing (CN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Dec. 28, 2022, as Appl. No. 18/089,750.
Int. Cl. G06F 9/50 (2006.01); G06F 9/38 (2018.01)
CPC G06F 9/5033 (2013.01) [G06F 9/3836 (2013.01); G06F 9/505 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method, in a data processing system, for scheduling workloads in a cloud computing system, the method comprising:
training a cloud affinity factor (CAF) computer model, via a machine learning process based on a training dataset comprising static characteristics of a workload binary for a workload, and dynamic characteristics corresponding to historical performance data for the workload, such that the trained CAF computer model predicts a first performance classification for a given workload binary;
processing, by the trained CAF computer model, a new workload to generate a second performance classification for the new workload;
generating one or more cloud affinity factors based on the second performance classification for the new workload;
applying at least one node affinity and dispatch rule to the one or more cloud affinity factors to select one or more nodes of the cloud computing system to which to dispatch the new workload; and
scheduling the new workload on the selected one or more nodes.